Evidence clearly indicates that abnormalities of blood coagulation and platelet function are present in overt diabetes mellitus; the contribution of these abnormalities, relative to genetic factors and abnormal lipid, carbohydrate, and amino acid metabolism to the development of microangiopathy and atherosclerosis, is, however, less clear. "Tight" diabetic control is not a universally accepted therapeutic principle, and short-term experiments to test whether important diabetes-related phenomena are reversible by tight control are desirable. In this study 10-20 diabetic subjects, preselected for the presence of clotting abnormalities, will be studied before and at the conclusion of a 3-6 week outpatient period of tight control achieved with a constant infusion insulin pump. Changes in platelet aggregation (ADP, collagen), plasma beta-thromboglobulin, platelet factor 4, platelet derived growth factor, thromboxane B2, fibrinopeptide A, and ristocetin cofactor will be correlated to improvement in diabetic control (HgbAIa-c, 24 hour urine glucose, fasting and post-prandial plasma glucose). Reversal of abnormalities would provide support for further studies to test whether long-range tight control can decrease the incidence of vascular and thrombotic complications, while a negative result will mean either that reversal might only be achieved by longer range control, or that the reversibility by tight control alone may not be possible in individuals who already have extensive blood vessel disease.